Posted on 30 August 2016
The study, shows for the first time that radiologists can detect breast cancer when viewing a mammogram of the opposite breast, even when there is no visible abnormal lesion.
This points to there being a ‘global signal’ not dependent on obvious visual markers that gives radiologists a ‘gist’ perception of whether cancer is present.
Studying radiologists at NHS Hospital Trusts in Yorkshire and Cambria, UK and U. T. MD Anderson Cancer Center in Houston, Texas, USA, researchers found that breast density levels or symmetry are not essential in forming a ‘signal’, but viewing the fine detail in a mammogram is crucial in making a ‘gist’ decision.
Characterising this signal is useful in that it could allow potential algorithms, based on perceptual analysis, to be developed to improve CAD systems. Such knowledge can also be incorporated into training protocols for medical experts, improving cancer detection.
Dr Karla Evans, Lecturer at York’s Department of Psychology and lead author of the study, said: “This study proves that radiologists can identify breast cancer at a glance, detecting subtle abnormalities using visual information that isn’t localised to a specific marker. This ‘global signal’ uses global image statistics to non-selectively process the whole mammogram image, rather than zoning in on a particular local indicator.
“This general ‘gist’ is important, particularly in cases where screening can produce a false abnormality. The team found that radiologists could do better than chance in discriminating breast cancer cases from normal tissue, even when the images of abnormal breast tissue did not directly capture a cancerous lesion or when those images were taken from the contralateral breast (the breast on the other side of the body) of a woman with breast cancer.
“Studies show that false negative screening results can often later develop into cancer, but by combining this ‘gist’ perception with medical screening we can monitor such patients more thoroughly.
“Radiologists often have ‘hunches’ about images on first glimpse. Our work shows that those hunches are based on something real in the image. Looking to the future, any medical field that requires image screening and diagnostics, such as dermatologists and pathologists, we think might use analogous signals.”